Local and landscape factor influences on avian community composition in playas of the Southern High Plains
MetadataShow full item record
The more than 25,000 playas on the Southern High Plains (SHP) are essential to the maintenance of regional biodiversity. The SHP landscape has been highly fragmented by intensive cultivation since the 1930s. Cultivation has degraded the condition of playas through sedimentation, resulting in a loss of playa volume in cropland watersheds and shortened hydroperiod. These effects should influence the avian communities that depend on playas for breeding, overwintering, or migratory stopover sites. However, information about the influence of land-use change and other factors on avian community composition in SHP playas is limited. Previous studies have focused on the relationship between game species and local variables during winter; however, systematic documentation of playa variables that are important to birds throughout the year is lacking. To fully understand the importance of playas to different avian communities, information on avian use during the wet and dry phases of playas throughout the year is needed. Therefore, I examined influences of local and landscape factors on avian species richness, diversity, and density in playas and determined how avian community composition was influenced by water-level fluctuation as a disturbance. My objectives were to 1) determine how local variables within a playa (e.g., vegetation cover and playa area) influence richness, density, and diversity of avian communities; 2) evaluate the effect of landscape variables in the wetland complex (i.e., playas and the adjoining habitat patches in 3 defined diameters) on richness, density, and diversity of avian communities; 3) examine the effect of water-level fluctuations as a disturbance on avian community composition; and 4) examine relationships between local variables and percent composition of different avian groups and dominant species during summer. After significant precipitation events in June 2003 and 2004, I selected 80 playas (40 playas each year) across the SHP containing water with watersheds ranging from 100% native grassland to 100% cropland. Wet playas were surveyed biweekly by counting all birds present within the playa boundary for as long as playas retained water, from June 2003 to May 2004 and June 2004 to May 2005. When playas dried, I conducted surveys monthly using distance sampling techniques. Water depth was measured concurrently with each bird survey and percent vegetation cover, vegetation structure, and plant species richness were measured 3 times during each summer. Eight landscape variables (i.e., number of playas, percent cover of playas, interspersion and juxtaposition index, landscape shape index, Shannon evenness index, diversity index of land use, edge density, and mean edge density) were obtained using ESRI® ArcGIS and FRAGSTATS*ARC® for each study playa at 3 distance scales (i.e., 1, 5, and 10 km radii from the study playas, representing a gradient of local to landscape-level influences). I recorded 127 bird species during 1487 surveys over 2 years. Seventy-seven were wetland-dependent species and 50 were non-wetland species. In wet playas, area was a positive predictor for total species richness and diversity in all seasons. Water depth was a consistent predictor across different seasons for total species richness, and species richness was greatest at an intermediate level of water depth (ranging from 40 to 100 cm). Tilled index ([tilled landscape ¡V untilled landscape]/[tilled landscape + untilled landscape]) positively influenced total species richness and native species richness, but I also found a positive relationship between percentage of exotic bird species in the avian community and tilled index, indicating that more exotic bird species (fewer native species) were found in playas surrounded by a greater percentage of cultivated land. Models predicting avian species richness were variable among seasons, and diversity of land use within 1 km had a positive influence on total species composition. In general, variables in the best-fit models for wetland-dependent species richness, diversity, and density were similar to models predicting total species. Models predicting species richness, diversity, and bird density for wetland-dependent species explained more variation than predictive models for total species richness. This is likely because wetland-dependent species outnumbered non-wetland species and were important in driving avian community composition in wet playas. Playa area had a positive influence on waterfowl species richness and density, whereas water depth had a polynomial relationship with waterfowl species richness and density, showing a positive linear and a negative quadratic term. Waterfowl species richness and density peaked when water depth ranged from 40-100 cm. Percent vegetation cover had a polynomial relationship with waterfowl species richness, indicating that waterfowl most frequently occupied wetlands with intermediate levels of vegetation. Landscape variables were important in predicting waterfowl species richness and density, but the appearance of landscape variables in best-fit models varied among seasons, which suggests different habitat requirements for waterfowl in different portions of the annual cycle. Playa area was positively correlated with shorebird species richness but there was a negative relationship between shorebird density and playa area. There was a polynomial relationship, showing a positive linear and a negative quadratic term between water depth and shorebird species richness and density, which indicates that shorebirds tended to occupy wetlands with intermediate levels of water depth (30-80 cm). Percent vegetation cover was negatively correlated with shorebird density. Playa area was a consistent positive variable for predicting wading bird species richness and density. Because of high evaporation rates in summer and unpredictable precipitation, playa hydroperiods are unpredictable and the number of playas that contain water varies among seasons and years. Therefore, the natural wet-dry cycle is ideal for studying the influences of disturbance on wetland biota. The intermediate disturbance hypothesis (IDH) states that the highest levels of species diversity will occur at intermediate levels of disturbance, whereas diversity will be lower at higher and lower levels of disturbance, owing to effects of abiotic limitations and competitive exclusion, respectively. I tested the predictions of the IDH by comparing species richness and similarity index (changes of avian composition; Sorenson¡¦s similarity coefficient) among 3 disturbance regimes (considered different combinations of disturbance intensity and frequency and starting moisture condition). Results did not support the IDH, however, as the highest species richness did not occur at intermediate levels of disturbance. These results suggest that moisture condition itself may be more important in influencing species richness than a change in moisture condition (i.e., disturbance). Because wet playas have more species than dry playas in this semi arid landscape, playas associated with flooded conditions may have higher species richness regardless of the degree of disturbance. Playas with a higher degree of disturbance were associated with higher tilled indices in summer. The increased degree of disturbance due to sedimentation and lack of protection from the cultivated watershed had a negative impact on avian community composition. I used canonical correspondence analysis to examine the influence of local environmental variables (i.e., playa area, tilled index, water depth, percent vegetation cover, vegetation structure, vegetation height) on percent composition of the avian assemblage. Waterfowl, shorebirds, and wading bird were analyzed as groups, along with individual analyses on blue-winged teal (Anas discors), mallard (Anas platyrhynchos), American avocet (Recurvirostra americana), killdeer (Charadrius vociferus), long-billed curlew (Numenius americana), upland sandpiper (Bartramia longicauda), mourning dove (Zenaida macroura), and red-winged blackbird (Agelaius phoeniceus). Water depth, vegetation structure, and percent vegetation cover were consistently the top 3 variables influencing percent composition of birds throughout the summer. In general, percent composition of waterfowl as a group peaked when water depth ranged from 40-70 cm. Percent composition of shorebirds as a group was negatively correlated to water depth and vegetation-related variables, indicating that shorebirds occupied playas with shallower water and sparse vegetation. Percent composition of long-billed curlew was negatively correlated with tilled index, suggesting that playas within grassland watersheds are important for this species. In contrast, percent composition of mourning dove and red-winged blackbirds were positively associated with vegetation structure and percent vegetation cover. Watershed management surrounding playas should be given a high priority depending on the goals in the conservation plans, as increasing cultivation in playa watersheds increases total species richness but also favors exotic avian species and increases the degree of disturbance (i.e., water level fluctuation). With shortened hydroperiod of sedimented playas, the value of playas for avian species is decreased. Larger playas should also receive higher emphasis when other conditions (e.g., land use, vegetation, and landscape variables) are the same, since conserving larger playas is likely to maximize species richness. At a local scale, maintaining an intermediate level of water depth (40-100 cm) and 26-50% of vegetation is likely to meet the needs for the highest number of species and for most of the bird groups. Variables at a landscape level should be considered in conservation plans, as landscape-level variables appeared in many of the top models. Considering a group of playas as a complex instead of dealing with individual playas should provide diverse habitat for different groups of birds.